Description

Data from a low-density polyethylene production process. There are 14 process variables and 5 quality variables.

Dataset

This is simulated data. Two tubular reactors are connected in series. The first 50 observations are from common-cause (normal) operation, while the last 4 show a process fault developing: the impurity level in the ethylene feed in both zones is increasing. The dataset consists of:

• Tin : inlet temperature to zone 1 of the reactor [K]
• Tmax1 : maximum temperature along zone 1 [K]
• Tout1 : outlet temperature from zone 1 [K]
• Tmax2 : maximum temperature along zone 2 [K]
• Tout2 : outlet temperature from zone 2 [K]
• Tcin1 : temperature of inlet coolant to zone [K]
• Tcin2 : temperature of inlet coolant to zone 2 [K]
• z1 : percentage along zone 1 where Tmax1 occurs [%]
• z2 : percentage along zone 2 where Tmax2 occurs [%]
• Fi1 : flow rate of initiators to zone 1 [g/s]
• Fi2 : flow rate of initiators to zone 2 [g/s]
• Fs1 : flow rate of solvent to zone 1 [% of ethylene]
• Fs2 : flow rate of solvent to zone 2 [% of ethylene]
• Press : pressure in the reactor [atm]
• Conv : quality variable: cumulative conversion
• Mn : quality variable: number average molecular weight
• Mw : quality variable: weight average molecular weight
• LCB : quality variable: long chain branching per 1000 C atoms
• SCB : quality variable: short chain branching per 1000 C atoms

Preparation

Python version: 3.6

License

Please follow this for License information